Diabetes is a complex, multi-factorial condition, affecting 26 million people in the United States and an additional 79 million people are at high-risk of developing diabetes. However, 90% of people at high-risk of diabetes don't know about their risk and very few are given guidance by their health care providers about improving their diets, physical activity, and lowering weight to reduce their risk of getting diabetes. We also know very little about how much these behavior choices are influenced by one's social connections, particularly partners or spouses that live in the same household, share resources, and have the same values and access to similar information. For example, emergence of diseases like cancers or diabetes can heighten family members' concerns of also developing disease, and stimulate them to engage in preventive behaviors. We aim to investigate these social spillover effects on preventive behaviors and diabetes risk within households by analyzing longitudinal data from 2005 to 2014 from a large, diverse insured population. Kaiser Permanente Northern California has a longitudinal and multi-ethnic database of 3.3 million members and offers a rich opportunity to link biometric, laboratory, and health service utilization data for primary plan members and their linked dependents. We will study whether a new diabetes diagnosis in the household prompts partners or spouses to change their lifestyle behaviors; whether this adoption of preventive behaviors is different across demographic or socioeconomic groups; and whether these behaviors are maintained 3-6 years after the diagnosis. We will also study whether adult family members that are on the same health insurance plan and live in the same household of someone that was newly diagnosed with diabetes in this time have a high risk of getting diabetes themselves. And lastly, whether the lifestyle behavior changes adopted are associated with reduced risk of diabetes in the family member? The findings from this study may lead to new ways of identifying people at high risk of diabetes - for example, doctors notifying people with new diabetes diagnoses to encourage their household members to get checked too. In addition, it may also lead to new studies to help whole families change and maintain their lifestyle behaviors together. Our proposal aims to use a unique and underutilized set of databases to answer these topical and innovative questions. Utilizing the combined and interdisciplinary expertise of our Emory University and KPNC study team, we will use advanced statistical and quasi-experimental methods to analyze the data, controlling for measurable biases and using sensitivity analyses to explore factors that cannot be accounted for. This study is novel in concept, efficient in design, and findings will have important implications for screening recommendations, for advising people with a new diagnosis of diabetes, and for exploring avenues to stimulate sustainable healthful behaviors through families.